An 87% efficient Spam Filter implemented from scratch using Naive Bayes Algorithm.
We coded a Naive Bayes classification model which is based on multinomial Naive Bayes algorithm. An accuracy of 87% was achieved on a test dataset, which is relatively good! It can be used to classify new messages as either spam or non-spam or ham.
We used the dataset of 5,572 SMS messages that are already classified or labelled by humans. The dataset was put together by Tiago A. Almeida and José María Gómez Hidalgo, and it can be downloaded from the The UCI Machine Learning Repository or from Amazon.